CN104192144A - Automobile active anti-collision curve false-alarm eliminating method - Google Patents

Abstract

The invention relates to an automobile active anti-collision curve false-alarm eliminating method. According to the automobile active anti-collision curve false-alarm eliminating method, an automobile is idealized into a mass point, and the automobile runs along a central line of a curve; an azimuth angle is defined to be the included angle between the straight line formed by two automobiles and the straight line formed by the exact front of the automobile, and the changing range of the azimuth angle in the same lane is worked out along the dynamic changes of the present range D of the two automobiles. The relative distance D and the azimuth angle theta are measured through a video or a millimeter-wave radar or a laser radar sensor, whether the azimuth angle is within the range or not is judged, and if yes, it is indicated that the two automobiles are in the same lane; if not, the two automobiles are not in the same lane, and a false alarm can be eliminated. The automobile active anti-collision curve false-alarm eliminating method has great practical significance and wide application prospects for solving the problem of traffic safety, improving transportation capability, reducing the occurrence rate of terrible traffic accidents, reducing losses of life and properties and improving social and economic benefits.

Description

Initiatively crashproof bend false-alarm removing method of a kind of automobile

Technical field

The invention belongs to automobile technical field, relate to a kind of autocontrol method for automobile, relate in particular to a kind of for the initiatively method of the elimination bend false-alarm of anti-collision technique of automobile.

Background technology

As the advanced vehicle of modernization, automobile has changed people's life style, has promoted the progress of socioeconomic development and human culture, brings greatly easily simultaneously to people's life, has also brought serious traffic safety problem.In order to reduce traffic accident and personal casualty, each state, all in positive studying the countermeasure, utilizes the whole bag of tricks and measure to reduce the generation of traffic accident.Moreover, the developing direction in anti-collision system for automobile and automobile future is closely related, and in not far future, it is simple and convenient that vehicle drive is bound to become, the dependence of the driving technique level height to personnel is bound to become more and more lower, until realize complete automatic Pilot.And to realize automatic Pilot, and automobile must possess failure-free collision avoidance system, particularly initiatively crashproof, and this is precondition and the important leverage of safe driving, is the first step of moving towards this long march of ten thousand li of automatic Pilot technology.

Due to the developing by leaps and bounds of Eltec, correlation technique is maked rapid progress in recent years, especially developing rapidly of information industry, makes collision-proof alarm and self-actuating brake control technology become possibility.Anti-collision system for automobile is mainly divided into the passive crashproof and crashproof two kinds of modes of active, and the former is alarm device, reminds and main chaufeur control the abrupt deceleration vehicle of relying on; The latter is except possessing chaufeur prompting function in the time that dangerous situation occurs, if not braking in time of chaufeur, autonomous cruise speed system can start dead-man's device, until remove the dangerous condition.

The greatest problem that automobile collision preventing is controlled at technical existence is false-alarm problem, and in this external bend and road slope situation, how the correct front risk object that identifies is also the problem that research automobile collision preventing technology must be considered.The effect of front collision warning is the safety hidden danger existing in alerting drivers road ahead traffic environment, reminds chaufeur to take measures in advance, the generation avoiding traffic accident.Generally, the vehicle travelling on this front side road on express highway is the principal element that forms this car safety hidden danger.Forwardly in collision warning algorithm, straight way section is because the well-regulated geometric configuration of tool is easy to differentiate the relative position relation of different automobiles on the different straight ways in front, find out and be positioned at the obstacle vehicle of this track to this car formation the greatest danger, thereby set up front alarm algorithm.And bend section relies on merely radar cannot realize the judgement to vehicle relative position relation on different tracks, impalpable forms dangerous vehicle to this car in this track, often occurs more false-alarm.

Summary of the invention

The present invention is directed to the deficiencies in the prior art, initiatively crashproof bend false-alarm removing method of a kind of automobile is provided.

Technical scheme of the present invention is:

This car is idealized as to particle, and this car travels along bend middle line; Definition azimuth, for straight line and this car dead ahead that two cars form form the angle between straight line, along with the dynamic change of two car present range D, calculates the scope that azimuth changes in same track:

$\mathrm{\θ}\∈(\frac{\mathrm{\π}}{2}-\arccos \frac{4D^2-L^2-4\mathrm{RL}}{8\mathrm{DR}},\frac{\mathrm{\π}}{2}-\arccos \frac{4D^2-L^2+4\mathrm{RL}}{8\mathrm{DR}})$

Wherein R is road radius in express highway, L Wei Nei road width, and θ is azimuth;

Measure two car relative distance D and azimuth angle theta by video, millimeter wave radar or laser radar sensor, judge whether azimuth is dropped in above-mentioned scope, if so, illustrate that two cars are in same track; Otherwise not in same track, this can reject for false-alarm;

In described express highway, the computation process of road radius R is:

The motion of automobile is interpreted as to a Motion of Rigid Body on road plane, and what the mode of motion of automobile was the translation of automobile self and automobile around the rotation of its barycenter is synthetic; Suppose that predicted time is T _p, based on the hypothesis to chaufeur forward sight strategy, for the current kinematic parameter t moment horizontal stroke of each given automobile, longitudinal acceleration and horizontal stroke, longitudinal velocity, taking current automobile position as coordinate origin, predict next moment t+T _pthe position of automobile obtains speed v, acceleration/accel a and course angle α under this desired location simultaneously.

According to infinitely small principle, by T _pbe divided into J equal portions, for each small equal portions Δ t, obtain respectively automobile longitudinal and horizontal state after this period; Be added to like this final step, obtain automobile at predicted time T _pafter state, the coordinate when obtaining automobile simultaneously and being positioned at this and the course angle of automobile, as the basis of next step calculating; The system of axes of the calculating institute reference of each step is different, must be transformed under the same coordinate system and just may be added; Choose system of axes (x ₀, y ₀) as the reference system of whole computation process, the i.e. frame of reference taking the bodywork reference frame of automobile current time as whole computation process, obtain this car residing position after predicted time, compare and determine this car safety whether after predicted time with the relative position of this position and front obstacle vehicle;

The speed of supposing each small moment i is v _x,i, v _y,i, acceleration/accel is a _x,i, a _y,i, course angle is α _i; If transition matrix is:

$A_i=\left[\begin{array}{cc}\cos {\mathrm{\α}}_i& -\sin {\mathrm{\α}}_i\\ \sin {\mathrm{\α}}_i& {\cos \mathrm{\α}}_i\end{array}\right]$

Thereby this car in each small moment coordinate position has formula:

$\left[\begin{array}{c}{x}_i\\ y_i\end{array}\right]=\left[\begin{array}{c}{x}_{i-1}\\ y_{i-1}\end{array}\right]+\left[\begin{array}{c}{v}_{x,i-1}\\ v_{y,i-1}\end{array}\right]\·\mathrm{\Δt}+A_{i-1}\·\left(\begin{array}{c}\frac{1}{2}\·\left[\begin{array}{c}{a}_{x,i}\\ a_{y,i}\end{array}\right]\·\mathrm{\Δ}{t}^2\end{array}\right)$

The coordinate position in above-mentioned each small moment is coupled together, obtain the imaginary line of this wheel paths; Adopt cubic algebra fitting of a polynomial automobile to expect the track travelling, and calculate thus the boundary curve of corresponding road both sides,

P ₃(x)＝a ₀+a ₁x+a ₂x ²+a ₃x ³

According to prediction locus data point, application method of least square, determines parameter a ₀, a ₁, a ₂, a ₃, can simulate prediction moment T _pwheelpath; In like manner, obtain the track that this car is expected traveling within following a period of time; Calculate the curvature of curve of this track, accordingly value Wei Nei road radius R.

The present invention concerning transport solution safely, improve transport capacity, reduce pernicious traffic accident incidence, reduce life and property loss and improve economic results in society, there is great realistic meaning and wide application prospect.

Brief description of the drawings

Fig. 1 is this car of bend false-alarm of the present invention and front truck orientation schematic diagram;

Fig. 2 is virtual road matching schematic diagram of the present invention;

Fig. 3 is the device block diagram that in active collision-avoidance system of the present invention, bend false-alarm is eliminated.

In figure, 1. this car, 2. front truck, 3. track bend, 4. this car fantasy sport track, 5. virtual road border, 6. range finding, angular measurement sensor, 7. car speed sensor, 8. steering angle sensor, 9. virtual road fitting algorithm unit, 10. active anti-collision control system false-alarm judging unit.

Detailed description of the invention

Below in conjunction with accompanying drawing, the present invention will be further described.

The present invention adopts azimuth observation method combined with virtual road fitting technology to reject false-alarm.Defined azimuth is that straight line and this car dead ahead that two cars form forms the angle between straight line, as shown in Figure 1, this car 1 and front truck 2 travel on bend 3, according to its speed of a motor vehicle and spacing, if according to Through Lane judgment rule, will send anti-collision warning or the Anticollision Measures of taking the initiative; But on its different tracks on bend 3, expection can not produce collision, therefore belongs to " false-alarm ".In order to remove this false-alarm, will judge that whether two cars are in same track, if be not just false-alarm in same track, if be exactly false-alarm in different tracks.Judge two cars whether in same track according to being exactly formula (1), whether drop on region shown in formula (1) according to two car azimuths and can judge.

In accompanying drawing 1, R is road radius in express highway, L Wei Nei road width, and D is two car present ranges, θ is azimuth.In order to simplify and computing without loss of generality, car is idealized as to particle, and this car travels along this track centre.With the dynamic change of D, can calculate the scope that azimuth changes in same track and be

$\mathrm{\θ}\∈(\frac{\mathrm{\π}}{2}-\arccos \frac{4D^2-L^2-4\mathrm{RL}}{8\mathrm{DR}},\frac{\mathrm{\π}}{2}-\arccos \frac{4D^2-L^2+4\mathrm{RL}}{8\mathrm{DR}})---\left(1\right)$

Measure two car relative distance D and azimuth angle theta by video, millimeter wave radar or laser radar sensor, judge whether azimuth is dropped in above-mentioned scope, if so, illustrate that two cars are in same track; Otherwise not in same track, this can reject for false-alarm.Wherein, in express highway, road radius R He Nei road width L is provided by virtual road fitting algorithm.

The method of the road boundary of traval trace and highway being carried out to matching is discussed below.From geometric angle, in real road environment, border, road left and right can be understood as the synthetic of curve and straight line.For this situation, the present invention is that the end to end little line segment in the border, left and right of road ahead is carried out curve fitting to the processing of road information, and like this, the boundary profile of road ahead is made up of many line segments in fact exactly, as shown in Figure 2.Relative distance between road boundary left and right wheels profile has been described the width L of road, and the slope variation (angle of this line segment and road transverse axis being defined as to the direction angle of road) of outline line has reflected that the curvature of road changes.Automobile route is predicted in variation according to road curvature, and therefore, automobile is taken aim in advance the variation that track has directly reflected road curvature within a period of time in future.

Under the prerequisite of the horizontal hypothesis of road, the motion of automobile can be understood as a Motion of Rigid Body on road plane, and what the mode of motion of automobile was the translation of automobile self and automobile around the rotation of its barycenter is synthetic.Suppose that predicted time is T _p, based on the hypothesis to chaufeur forward sight strategy, for the current kinematic parameter t moment horizontal stroke of each given automobile, longitudinal acceleration and horizontal stroke, longitudinal velocity, can, taking current automobile position as coordinate origin, predict next moment (t+T _pmoment) position of automobile, obtain the state parameters such as speed v, acceleration/accel a and course angle α under this desired location simultaneously.

According to infinitely small principle, by T _pbe divided into J equal portions, for each small equal portions Δ t, because time length is very short, can ignore automobile influencing each other between horizontal and vertical, therefore can utilize formula of reduction to obtain respectively automobile longitudinal and the horizontal state after during this period of time.Be added to according to the method final step, can obtain automobile and take aim in advance time T _pafter state, the coordinate when obtaining automobile simultaneously and being positioned at this and the course angle of automobile, as the basis of next step calculating.The system of axes of the calculating institute reference of each step is different, can not merely each step result of calculation be added, and must be transformed under the same coordinate system and just may be added.Choose system of axes (x ₀, y ₀) as the reference system of whole computation process, the i.e. frame of reference taking the bodywork reference frame of automobile current time as whole computation process, can obtain this car residing position after taking aim at the time in advance, compare and determine this car safety whether after taking aim at the time in advance with the relative position of this position and front obstacle vehicle.

The speed of supposing each small moment i is v _x,i, v _y,i, acceleration/accel is a _x,i, a _y,i, course angle is α _i.If transition matrix is:

$A_i=\left[\begin{array}{cc}\cos {\mathrm{\α}}_i& -\sin {\mathrm{\α}}_i\\ \sin {\mathrm{\α}}_i& {\cos \mathrm{\α}}_i\end{array}\right]---\left(2\right)$

Thereby this car in each small moment coordinate position has formula:

$\left[\begin{array}{c}{x}_i\\ y_i\end{array}\right]=\left[\begin{array}{c}{x}_{i-1}\\ y_{i-1}\end{array}\right]+\left[\begin{array}{c}{v}_{x,i-1}\\ v_{y,i-1}\end{array}\right]\·\mathrm{\Δt}+A_{i-1}\·\left(\begin{array}{c}\frac{1}{2}\·\left[\begin{array}{c}{a}_{x,i}\\ a_{y,i}\end{array}\right]\·\mathrm{\Δ}{t}^2\end{array}\right)---\left(3\right)$

The coordinate position in above-mentioned each small moment is coupled together, obtain the imaginary line of this wheel paths.Adopt cubic algebra fitting of a polynomial automobile to expect the track travelling, and calculate thus the boundary curve of corresponding road both sides,

P ₃(x)＝a ₀+a ₁x+a ₂x ²+a ₃x ³ (4)

According to prediction locus data point, application method of least square, determines the parameter a in formula (4) ₀, a ₁, a ₂, a ₃, can simulate following moment T _pwheelpath.In like manner, can obtain the track that this car is expected traveling within following a period of time.Calculate the curvature of curve of this track, value is bend road radius R accordingly, interior road width L is provided by lane mark identification algorithm or is given by standard vehicle width, substitution formula (1) can calculate azimuth variation range in same track, judge thus whether gained azimuth drops on above-mentioned scope, thereby false-alarm all can be rejected.

Shown in Fig. 3, two car present range D and azimuth angle theta by finding range, angular measurement sensor 6 obtains, range finding, angular measurement sensor 6 can be video frequency pick-up head, millimeter wave radar or laser radar etc.Interior road width L is provided by lane mark identification algorithm or is given by standard vehicle width; In express highway, road radius R is calculated by virtual road fitting algorithm unit 9.Obtained the state parameter such as moving velocity v, acceleration/accel a and course angle α of this car 1 by car speed sensor 7 and steering angle sensor 8, and calculate road radius R in express highway by virtual road fitting algorithm unit 9, finally by whether false-alarm of active anti-collision control system false-alarm judging unit 10 computational discrimination front trucks 2.

Claims (1)

1. an initiatively crashproof bend false-alarm removing method of automobile, is characterized in that:

This car is idealized as to particle, and this car travels along bend middle line; Definition azimuth, for straight line and this car dead ahead that two cars form form the angle between straight line, along with the dynamic change of two car present range D, calculates the scope that azimuth changes in same track:

$\mathrm{\θ}\∈(\frac{\mathrm{\π}}{2}-\arccos \frac{4D^2-L^2-4\mathrm{RL}}{8\mathrm{DR}},\frac{\mathrm{\π}}{2}-\arccos \frac{4D^2-L^2+4\mathrm{RL}}{8\mathrm{DR}})$

Wherein R is road radius in express highway, L Wei Nei road width, and θ is azimuth;

Measure two car relative distance D and azimuth angle theta by video, millimeter wave radar or laser radar sensor, judge whether azimuth is dropped in above-mentioned scope, if so, illustrate that two cars are in same track; Otherwise not in same track, this can reject for false-alarm;

In described express highway, the computation process of road radius R is:

The motion of automobile is interpreted as to a Motion of Rigid Body on road plane, and what the mode of motion of automobile was the translation of automobile self and automobile around the rotation of its barycenter is synthetic; Suppose that predicted time is T _p, based on the hypothesis to chaufeur forward sight strategy, for the current kinematic parameter t moment horizontal stroke of each given automobile, longitudinal acceleration and horizontal stroke, longitudinal velocity, taking current automobile position as coordinate origin, predict next moment t+T _pthe position of automobile obtains speed v, acceleration/accel a and course angle α under this desired location simultaneously;

According to infinitely small principle, by T _pbe divided into J equal portions, for each small equal portions Δ t, obtain respectively automobile longitudinal and horizontal state after this period; Be added to like this final step, obtain automobile at predicted time T _pafter state, the coordinate when obtaining automobile simultaneously and being positioned at this and the course angle of automobile, as the basis of next step calculating; The system of axes of the calculating institute reference of each step is different, must be transformed under the same coordinate system and just may be added; Choose system of axes (x ₀, y ₀) as the reference system of whole computation process, the i.e. frame of reference taking the bodywork reference frame of automobile current time as whole computation process, obtain this car residing position after predicted time, compare and determine this car safety whether after predicted time with the relative position of this position and front obstacle vehicle;

The speed of supposing each small moment i is v _x,i, v _y,i, acceleration/accel is a _x,i, a _y,i, course angle is α _i; If transition matrix is:

$A_i=\left[\begin{array}{cc}\cos {\mathrm{\α}}_i& -\sin {\mathrm{\α}}_i\\ \sin {\mathrm{\α}}_i& {\cos \mathrm{\α}}_i\end{array}\right]$

Thereby this car in each small moment coordinate position has formula:

$\left[\begin{array}{c}{x}_i\\ y_i\end{array}\right]=\left[\begin{array}{c}{x}_{i-1}\\ y_{i-1}\end{array}\right]+\left[\begin{array}{c}{v}_{x,i-1}\\ v_{y,i-1}\end{array}\right]\·\mathrm{\Δt}+A_{i-1}\·\left(\begin{array}{c}\frac{1}{2}\·\left[\begin{array}{c}{a}_{x,i}\\ a_{y,i}\end{array}\right]\·\mathrm{\Δ}{t}^2\end{array}\right)$

The coordinate position in above-mentioned each small moment is coupled together, obtain the imaginary line of this wheel paths; Adopt cubic algebra fitting of a polynomial automobile to expect the track travelling, and calculate thus the boundary curve of corresponding road both sides,

P ₃(x)＝a ₀+a ₁x+a ₂x ²+a ₃x ³

According to prediction locus data point, application method of least square, determines parameter a ₀, a ₁, a ₂, a ₃, can simulate prediction moment T _pwheelpath; In like manner, obtain the track that this car is expected traveling within following a period of time; Calculate the curvature of curve of this track, accordingly value Wei Nei road radius R.